Electrodeposition of Textured Ceramic Superlattices in the Pb−Tl−O

Superlattices in the Pb-Tl-O system were electrodeposited at room temperature from aqueous ... Cu2O/Cu nanocomposites,10 AgO,11 nanophase CeO2,12...
0 downloads 0 Views 396KB Size
1670

Chem. Mater. 1997, 9, 1670-1677

Electrodeposition of Textured Ceramic Superlattices in the Pb-Tl-O System Richard J. Phillips, Teresa D. Golden, Mark G. Shumsky, Eric W. Bohannan, and Jay A. Switzer* University of MissourisRolla, Department of Chemistry, and Graduate Center for Materials Research, Rolla, Missouri 65409-1170 Received February 5, 1997. Revised Manuscript Received May 5, 1997X

Superlattices in the Pb-Tl-O system were electrodeposited at room temperature from aqueous solution. The deposition solution was 0.005 M Tl(I) and 0.100 M Pb(II) in 5 M NaOH. The superlattices grew epitaxially on oriented prelayers with either a [100] or [210] texture. The out-of-plane texture of the prelayer was maintained throughout the growth of the superlattice. X-ray diffraction (XRD) showed the films to be superlattices by the presence of satellites around the main Bragg reflections. There was good agreement between XRD and scanning tunneling microscopy for the measurement of modulation wavelengths from 3 to 150 nm. Partial relaxation of coherency was observed for modulation wavelengths exceeding about 20 nm.

Introduction Superlattices are an example of a nonequilibrium structure that can be assembled electrochemically. The low processing temperature of electrodeposition should allow the growth of nanoscale architectures with minimal interdiffusion across interfaces. Electrodeposition also provides feedback during the deposition process, since the current during a potentiostatically controlled deposition provides a measure of the deposition rate. We have previously analyzed the current-time transients during the deposition of a superlattice to estimate the composition profile in the nanometer-scale layers.1 Most research on the electrodeposition of superlattices has been on metallic systems.2 The emphasis has been on growing nanoscale structures with enhanced elastic and plastic properties,3,4 improved wear resistance,5 and magnetic anisotropy.6 Schwarzacher has grown modulated nanoscale wires for giant magnetoresistance applications.7,8 Our interest lies in the deposition of nanometer-scale metal oxides with optical or electrical properties which are a function of the dimension of the material. We have shown that it is possible to electrodeposit Cu2O,9 Cu2O/Cu nanocomposites,10 AgO,11 nanophase CeO2,12 * To whom correspondence should be addressed. X Abstract published in Advance ACS Abstracts, June 15, 1997. (1) Switzer, J. A.; Phillips, R. J.; Golden T. D. Appl. Phys. Lett. 1995, 66, 819. (2) Moffat, T. P. J. Electrochem. Soc. 1995, 142, 3767. (3) Tench, D. M.; White, J. T. J. Electrochem. Soc. 1990, 137, 3061. (4) Lashmore, D. S.; Thomson, R. J. Mater. Res. 1992, 7, 2379. (5) Ruff, A. W.; Lashmore, D. S. Wear 1991, 151, 245. (6) McMichael, R. D.; Atzmony, U.; Beauchamp, C.; Bennett, L. H.; Swartzendruber, L. J.; Lashmore, D. S.; Romankiw, L. T. J. Magn. Magn. Mater. 1992, 113, 149. (7) Schwarzacher, W.; Lashmore, D. S. IEEE Trans. Magn. 1996, 32, 3133. (8) Attenborough, K.; Hart, R.; Schwarzacher, W.; Ansermet, J.-P. Mater. Res. Soc. Symp. Proc. 1995, 384, 3. (9) Golden, T. D.; Shumsky, M. G.; Zhou, Y.; VanderWerf, R. A.; Van Leeuwen, R. A.; Switzer, J. A. Chem. Mater. 1996, 8, 2499. (10) Switzer, J. A.; Hung, C.-J.; Bohannan, E. W.; Shumsky, M. G.; Golden, T. D.; VanAken, D. C. Adv. Mater. 1997, 9, 334. (11) Breyfogle, B. E.; Hung, C.-J.; Shumsky, M. G. Switzer, J. A. J. Electrochem. Soc. 1996, 143, 2741.

S0897-4756(97)00074-4 CCC: $14.00

and ZrO2,13 epitaxial Ag(Ag3O4)2NO3,14 compositional superlattices of Pb-Tl-O,15,16 and defect-chemistry superlattices based on Tl2O3.17 In this paper we provide details on the electrodeposition and characterization of superlattices in the Pb-Tl-O system with a strong outof-plane preferred orientation. Experimental Section Electrochemical experiments were performed using an EG&G Princeton Applied Research (PAR) model 273A potentiostat/galvanostat. A constant temperature was maintained at 25 °C by a Fisher Scientific Model 9100 circulator, and the solutions were stirred. The current response during superlattice deposition was recorded with a Nicolet Model 310 digital oscilloscope. The cell configuration consisted of the platinum counter electrode placed horizontally approximately 1 cm below a horizontal stainless steel working electrode. A standard calomel reference electrode within a bridge tube solution of 5 M NaNO3 was placed approximately 1 cm from the working electrode. A polyethylene frit was used at the end of the bridge tube. The working electrode was a 430 stainless steel disk (electrode area ) 1.96 or 3.88 cm2) supplied by Metal Samples, Inc. A platinum rotating disk electrode (area ) 0.138 cm2) was used as the working electrode for the linear sweep experiments. Contact was made to the electrode by attaching a copper wire to the back face of the electrode using silver epoxy available from Epoxy Technology, Inc. (415G). The electrode resistance was 0.5 Ω or less between the exposed face of the electrode and the copper wire. Once the silver epoxy was cured for 24 h, the electrode was mounted in Buehler epoxide resin (20-8130-128) and epoxide hardener (20-8132-032) in a 5:1 resin to hardener ratio, followed by a 24 h cure. The electrode was then polished by wet grinding through Buehler Carbimet 600 grit SiC paper. Final polishing was performed on Buehler Microcloth (40-7218) through 0.05 (12) Zhou, Y.; Phillips, R. J.; Switzer, J. A. J. Am. Ceram. Soc. 1995, 78, 981. (13) Switzer, J. A.; Phillips, R. J. Mater. Res. Soc. Symp. Proc. 1988, 121, 111. (14) Breyfogle, B. E.; Phillips, R. J.; Switzer, J. A. Chem. Mater. 1992, 4, 1356. (15) Switzer, J. A.; Shane, M. J.; Phillips, R. P. Science 1990, 247, 444. (16) Switzer, J. A.; Raffaelle, R. P.; Phillips, R. J.; Hung, C.-J.; Golden, T. D. Science 1992, 258, 1918. (17) Switzer, J. A.; Hung, C.-J.; Breyfogle, B. E.; Shumsky, M. G.; Van Leeuwen, R.; Golden, T. D. Science 1994, 264, 1505.

© 1997 American Chemical Society

Textured Ceramic Superlattices µm alumina. The electrode was rinsed with distilled water and ultrasonicated between each grinding/polishing stage. This epoxy mounting procedure produced a stainless steel disk flush with the surface of the epoxy, allowing more uniform mass transport control across the surface of the working electrode in a stirred solution. The Pb-Tl-O prelayers and superlattices were deposited from 0.005 M TlNO3 and 0.1 M Pb(NO3)2 in 5 M NaOH (caution: thallium and lead compounds are extremely toxic). Tl2O3 prelayers were deposited from 0.1 M TlNO3 in 5 M NaOH. The 5 M NaOH was prepared in deionized water. With this solution still warm, the nitrate salts were added, and the solution was allowed to cool and age for 24 h before deposition. The aging (while stirring) of the solution was needed to completely dissolve the salts in solution. The prelayers were grown to 20 and 5 µm for the Pb-Tl-O and Tl2O3, respectively. Subsequent superlattice growth on these prelayers did not result in a change of the texture throughout the thickness of the superlattice. The mixed oxide prelayers were deposited from a fresh stirred solution each time to minimize the effects of thallium ion depletion from the solution. All solutions were made from reagent grade chemicals dissolved in doubly distilled deionized water with a resistivity of at least 18 MΩ cm. Compositions of Pb-Tl-O films were analyzed with a JEOL JSM-T330A scanning electron microscope using energydispersive X-ray spectroscopy. The accelerating voltage was 20 keV with a 90° beam incident angle, 29.9° X-ray emergence angle, 0.5° X-ray window incidence angle, a channel to channel separation of 20 eV, and a data acquisition time of 100 s. The L-lines were used from the energy spectrum for analysis. X-ray diffraction patterns were run on a Scintag 2000 diffractometer using Cu KR radiation (λκR1 ) 0.154 059 8 nm). The tube source was operated at 45 kV and 40 mA. Two types of diffraction experiments were performed. The first was a symmetric scan used to evaluate the film texture and modulation wavelength perpendicular to the substrate. The second type of scan was used to determine the texture parallel to the substrate. Symmetric scans which were used to determine the overall texture utilized a step of 0.02° in 2 θ, with a 1 s dwell time per step. To obtain improved statistics around the satellite positions for the superlattice films, a step of 0.02° in 2 θ with a 10 s dwell time was used. The source slit widths were 1.0 mm and 2.0 mm. The detector was a Si-Li solidstate detector maintained at 77 K, with two receiving slits of 0.5 and 0.3 mm widths. Scans to evaluate the texture parallel to the substrate were run at a step of 0.1° and a 1 s dwell time per step. Scanning tunneling microscopy experiments were performed directly in air on cleaved cross sections of the Pb-Tl-O superlattice with a Digital Instruments Nanoscope II. Mechanically formed Pt/Ir tips were used as probes. All measurements were obtained with a bias voltage of 200 mV (sample positive), tunneling current of 0.5 nA, and scan rate of 4.34 Hz.

Results and Discussion Linear Sweep Voltammetry. The composition of an individual layer can be controlled during electrodeposition by having the solution concentration of the species which is oxidized at the lowest potential kept low and a separation between the standard reduction potentials of the individual components. For deposition of mixed metal Pb-Tl-O films, the [Tl(I)] is kept low in solution (5 mM), while [Pb(II)] is much higher (100 mM). The solution for mixed oxide deposition is compared to the individual solutions of 5 mM TlNO3 in 5 M NaOH and 0.1 M Pb(NO3)2 in 5 M NaOH using linear sweep voltammetry as shown in Figure 1. This shows that the open-circuit potentials (potential at zero current) of the individual solutions of Pb(II) and Tl(I) are separated by over 200 mV. Also, for a 5 mM Tl(I) concentration, a mass-transport-limited current density of 2.3 mA/cm2 was measured. The linear sweep voltammetry indicates that for the range of potentials used to

Chem. Mater., Vol. 9, No. 7, 1997 1671

Figure 1. Linear sweep voltammetry for individual and multiple component solutions. The supporting electrolyte was 5 M NaOH in a stirred solution. The scan rate was 10 mV/s.

deposit the superlattices, the Tl(I) deposition is masstransport-limited. Since a mass transport limited current is not observed for Pb(II), this component is deposited under activation or kinetic control. For the linear sweep curves in Figure 1, the sum of the currents from the individual solutions do not add up to the measured current for the mixed solution. The current from the mixed solution increases above the mass transport limited current for Tl(I) approximately 150 mV before the oxidation of Pb(II) was observed in the 0.1 M Pb(NO3)2 solution. Also, the current in the mixed solution appears to be suppressed, and the Tl(I) mass-transport-limited current obtained in the 5 mM TlNO3, 5 M NaOH solution is not observed in the mixed solution before 150 mV vs SCE. A possible explanation for the voltammetry results of the mixed Pb-Tl solution could be the result of adsorbed Pb(II) species. Fleischmann et al.18 proposed a mechanism for the deposition of PbO2, where adsorbed Pb(OH)+2,ads or Pb(OH)2+2,ads were involved in the rate-determining steps in the deposition. The adsorbed species may be responsible for the decreased current of the mixed solution before 150 mV vs SCE. A similar shifting of the deposition potentials for the mixed oxide was also observed by cyclic voltammetry.19 To examine the effect of Pb(II) on the deposition of Tl(I), we added varying amounts of Pb(NO3)2 to a 5 mM TlNO3, 5 M NaOH solution. Figure 2 shows the linear sweep voltammetry results for stirred solutions of 5 mM TlNO3 in 5 M NaOH with varying amounts of Pb(II) added to the Tl(I) solution. Curve A shows the linear sweep with no Pb(II) in solution. As the concentration of Pb(II) is increased, the current tends to rise more rapidly for a given applied potential (i.e., solutions B and C). A further increase in the Pb(II) concentration to 5 mM (i.e., solution D) results in a current rise similar to a solution without Pb(II) (i.e., solution A). An increase in the limiting current from 2.3 mA/cm2 for 0 mM Pb(II) to 4.7 mA/cm2 for 5 mM Pb(II) is observed. As the concentration of Pb(II) is increased to 75 mM (solution E) and 100 mM (solution F), a definitive Tl(I) limiting current is not observed, and the voltammogram (18) Fleischmann, M.; Thirsk, H. R. Electrochim. Acta 1960, 2, 22. (19) Voos, B.; Pleth, W.; Borner, R. H. K.; Schollhorn, R. Electrochim. Acta 1993, 38, 2551.

1672 Chem. Mater., Vol. 9, No. 7, 1997

Phillips et al. Table 2. Comparison of Bulk Film Compositions by Energy-Dispersive X-ray Spectroscopy (EDS) and Anodic Stripping Voltammetry (ASV) deposition current atomic atomic atomic atomic potential density percent percent percent percent (mV vs SCE) (mA/cm2) Pb (ASV) Tl (ASV) Pb (EDS) Tl (EDS) 70 150 230

0.05 1.0 5.0

54 68 79

46 32 21

45 64 76

55 36 24

Figure 2. Linear sweep voltammetry for a 5 mM TlNO3, 5 M NaOH stirred solution with (A) 0 mM, (B) 0.01 mM, (C) 1.0 mM, (D) 5.0 mM, (E) 75 mM, and (F) 100 mM Pb(NO3)2. Solution (G) is 100 mM Pb(NO3)2 in 5 M NaOH. The scan rate was 10 mV/s and the electrode area 0.138 cm2. Table 1. Summary of Open-Circuit Potentials for 1 µm Thick Tl2O3, PbxTlyOz, and PbO2 Films in a Solution of 5 MM TlNO3, 100 MM Pb(NO3)2, and 5 M NaOH applied potential (mV vs SCE) 80a 70b 120b 150b 230b 260b 295c

open-circuit potential (mV vs SCE) -15 -5 15 22 51 56 60

Figure 3. Dependence of the prelayer texture on the film thickness for a Tl2O3 prelayer. The texture increased with film thickness. A current density of 5 mA/cm2, and a stirred solution of 0.1 M TlNO3 in 5 M NaOH was used to deposit the prelayers.

atomic percentd Tl

Pb

100 55 42 36 24 18 0

0 45 58 64 76 82 100

a Deposited from 100 mM TlNO , 5 M NaOH, stirred. b Depos3 ited from 5 mM TlNO3, 100 mM Pb(NO3)2, 5 M NaOH, stirred. c Deposited from 100 mM Pb(NO ) , 5 M NaOH, stirred. d Deter3 2 mined on a metals-only basis.

shifts toward a solution containing only 100 mM Pb(II) (solution G). The open-circuit potentials were measured in a mixed solution of 5 mM TlNO3, 100 mM Pb(NO3)2, and 5 M NaOH for 1 µm thick Tl2O3, PbxTlyOz, and PbO2 films as shown in Table 1. These films were obtained by depositing a [100] Tl2O3 prelayer followed by deposition of a 1 µm film in their respective solutions using the applied potentials shown. For PbxTlyOz deposited at 70 mV vs SCE, the open circuit is similar to the measured open circuit for Tl2O3. As the lead content increases, the open circuit also increases, and films deposited at 260 mV vs SCE approach an open-circuit similar to a PbO2 film. This indicates that each of the compositions of the Pb-Tl-O mixed oxides have a characteristic open-circuit potential which is different from the pure components, due to the varying Tl/Pb ratio within the films. The linear sweep voltammetry and open circuit measurements for Pb-Tl-O films suggest that both kinetics and thermodynamics can affect the deposition from a single plating bath. Each composition should have a characteristic deposition potential. Consequently, it is difficult to calculate the film composition in advance using the current from the linear sweep scan

in a multiple component solution such as 5 mM TlNO3, 100 mM Pb(NO3)2 in 5 M NaOH. Instead, compositions for bulk films grown either galvanostatic or potentiostatic can be determined by X-ray spectroscopy or electrochemical analysis. The relative compositions of Pb-Tl-O films were first analyzed using energydispersive X-ray spectroscopy (EDS) in a scanning electron microscope (SEM). The EDS results were confirmed by anodic stripping voltammetry (ASV) for dissolved films using the method of standard additions. The Pb-Tl-O films were first dissolved in 50 mM ascorbic acid and the Pb(II) in solution was complexed with EDTA (C10H16N2O2) to shift the Pb(II) reduction peak to more cathodic potentials. The Tl(I) and Pb(II) were then preconcentrated from the solution into a mercury drop working electrode at -1.0 V versus Ag/ AgCl. Following the preconcentration step, the ions were stripped electrochemically by anodically scanning the potential. The ASV results are compared to the EDS results in Table 2. These results confirm that the composition within these bulk films can be controlled with the applied potential or current during deposition. Textured Prelayers. Prelayers of [210] oriented Pb-Tl-O and [100] oriented Tl2O3 were deposited onto stainless steel substrates at 5 mA/cm2. The Pb-Tl-O prelayers had a fluorite crystal structure with space group Fm3m, and a lattice parameter of 0.5342 nm. The Tl2O3 prelayers had a bixbyite crystal structure with space group Ia3, and a lattice parameter of 1.0543 nm. The bixbyite structure is often described as a doubled fluorite structure with one-quarter of the oxygens missing. The mismatch between the doubled Pb-Tl-O and the Tl2O3 structure is 1.3%. The evolution of texture for the prelayers is demonstrated in Figure 3 for a series of thallic oxide prelayers. By a thickness of 5 µm, Tl2O3 prelayers were mainly [100] oriented. For the Pb-Tl-O prelayer, the texture

Textured Ceramic Superlattices

Figure 4. Highly oriented 20 µm Pb-Tl-O prelayer with a [210] texture. A current density of 5 mA/cm2, and a stirred solution of 0.005 M TlNO3, 0.1 M Pb(NO3)2 in 5 M NaOH was used to deposit the Pb-Tl-O film.

also increased as the film thickness increased. A PbTl-O prelayer shown in Figure 4 has the (420) reflection as the main peak at a thickness of 20 µm. The high intensity of the (420) reflection relative to the other peaks in the X-ray spectrum is an indication of a strongly preferred orientation with the {210} planes parallel to the substrate surface. These textured prelayers were used as substrates for subsequent superlattice deposition. The grain size for the textured prelayers was measured parallel to the stainless steel substrates on the surface of the films using the line-intercept method in scanning electron microscopy (SEM). Figure 5A shows the variation of grain size with film thickness for Tl2O3 prelayer films. The grain size increased from 0.25 µm at a film thickness of 0.5 µm to approximately 2 µm at a film thickness of 5 µm. The microstructure of the 5 µm thick Tl2O3 film shown in Figure 5B consists of faceted grains. The Tl2O3 films appear dense on the surface for a 5 µm thickness. The average grain size for the Pb-Tl-O prelayer also increases with the film thickness as shown in Figure 6A. Initially, the grain size was approximately 0.5 µm for a film thickness of 1 µm and increased to over 2 µm for a 20 µm thick film. As the film thickness increases to 20 µm, the grains appear rounded without obvious faceting as shown in Figure 6B. We have characterized the oriented prelayers with X-ray diffraction using azimuthal rotation scans. These scans provide information on the texture parallel (i.e., in-plane) to the substrate.20 Azimuthal scans are obtained by tilting the substrate to an appropriate angle such that another set of planes, which are an allowable (20) Klug, H. P; Alexander, L. E. X-ray Diffraction Procedures for Polycrystalline and Amorphous Materials, 2nd ed.; John Wiley & Sons: New York, 1974.

Chem. Mater., Vol. 9, No. 7, 1997 1673

Figure 5. (A) Dependence of the average grain size with Tl2O3 film thickness. (B) Surface morphology of the [100] oriented oxide. The 5 µm thick film shows a faceted surface.

Figure 6. (A) Dependence of the average grain size with PbTl-O film thickness. (B) Surface morphology of the [210] oriented oxide with a film thickness of 20 µm.

reflection for that structure, are positioned parallel to the substrate (in-plane position). The {111} planes satisfy this requirement for the electrodeposited prelayer films. The tilt angles were 39.2° for the [210] oriented mixed oxide and 54.7° for the [100] oriented thallic oxide. After tilting, the samples are rotated by a 360° azimuthal angle to determine if in-plane orientation exists. The tube and detector are positioned at the Bragg condition for the {111} planes. Figure 7 compares in-plane textures for (A) [100] silicon single

1674 Chem. Mater., Vol. 9, No. 7, 1997

Phillips et al.

The microscopic or atomistic mechanisms which result in textured films are not universally understood. In general, the fastest growing direction approaches a perpendicular alignment with the original surface as the film growth proceeds. For electrodeposited films, the exchange current density, jo, has been shown to vary for different crystal faces.21 Deposition rates on different faces can be altered by adsorbed species such as hydrogen or organic additives. Also, a physically nonuniform surface can create a different current density at a specific point resulting in the alteration of the deposition rate. Electrodeposited films of Tl2O3 have a strong [100] texture, indicating the {100} planes have the fastest growth rate. In contrast, Wirtz and Siebert showed that the 〈111〉 directions were the fastest growing for bcc-Tl2O3 grown by vapor deposition.22 There are other examples of electrodeposited films which also exhibit a [110] texture. Electrodeposited Cu/Ni superlattices have been fabricated primarily with a [110] texture.23 These multilayers contain (110), (100), and (111) facets. The modulation wavelength was measured to be the largest on the (110) planes followed by the (100) and (111) planes. The texture of these electrodeposited films are in contrast to sputtered Fe/Pd superlattice films,24 where the closest packed planes grow the fastest and are parallel to the substrate [i.e., (110) for bcc-Fe and (111) for fcc-Pd]. Superlattice Growth. Superlattices electrodeposited onto these prelayers in the Pb-Tl-O system consist of alternating layers, where one layer is stacked on top of another without a change in the crystallographic orientation normal to the substrate. The superlattice structures can be deposited by pulsing either the current or potential. Composition profiles have been calculated for potential pulsed superlattices.1 Unless otherwise stated, the superlattices grown for this paper were made by pulsing the current density between 0.5 and 5.0 mA/ cm2. A schematic of a superlattice is shown in Figure 8. The individual layer thickness (λa or λb) in Figure 8 depend on the deposition times at the respective current densities for each layer. The thickness of layer a plus layer b equals the modulation wavelength, Λ, of the superlattice. For an electrochemically deposited film, the modulation wavelength, Λ, can be estimated from Faraday’s law:

moles )

1 Q ) nF nF

∫iti dt )

FAλF M

(1)

by rearranging eq 1 to Figure 7. Rotational scans for (A) [100] silicon single crystal, (B) silicon powder, and (C) a 5 µm thick [100] Tl2O3 film. X-ray diffraction is from the {111} planes.

λF ) MQ/nFAF

(2)

and substituting from the following equations: crystal, (B) silicon powder, and (C) a 5 µm thick Tl2O3 film. If in-plane texture exists, the symmetry for the {111} planes will be 4-fold, and the azimuthal scans will contain four maxima. This is demonstrated for singlecrystal silicon by Figure 7A. A random sample, such as silicon powder, would not have distinct maxima as shown in Figure 7B. As seen in Figure 7C, the in-plane texture of the electrodeposited film was also random. Similar results are obtained for mixed oxide prelayers and the superlattices deposited onto these prelayers.

ΛF ) λF,a + λF,b

(3)

na ) nb ) n

(4)

(21) Bockris, J. O’M.; Reddy, A. K. N. Modern Electrochemistry, 3rd ed.; Plenum Press: New York, 1977; Vol. 2. (22) Wirtz, G. P.; Siebert, D. C. J. Cryst. Growth 1976, 32, 274. (23) Haseeb, A. S. M. A.; Celis, J. P.; Roos, J. R. J. Electrochem. Soc. 1994, 141, 230. (24) Boufelfel, A.; Emrick, R. M. Phys. Rev. B 1991, 43, 13, 152.

Textured Ceramic Superlattices

Chem. Mater., Vol. 9, No. 7, 1997 1675

Figure 8. Idealized superlattice grown on an oriented prelayer. The orientation of the prelayer is maintained through the superlattice. The thickness of layer a is λa, and the thickness of layer b is λb. Figure 10. (A) Surface morphology of a [100] textured superlattice grown on [100] Tl2O3. The faceting from the prelayer was not observed on the surface of the superlattice. (B) Surface morphology of a [210] textured superlattice without the presence of facets. Both superlattices were grown by current pulsing. The thicknesses of the superlattices were 6 and 4.5 µm for the [100] and [210] textures, respectively.

Figure 9. Scanning electron micrograph of an electrodeposited columnar superlattice structure. The individual layer thicknesses are approximately 0.5 µm. The bar marker is 1.0 µm.

then the equation for the modulation wavelength is

ΛF )

(

)

1 QaMa QbMb + nFA Fa Fb

(5)

For galvanostatically deposited superlattices, the charge for the individual layer Qi, is replaced by the product (it) which results in:

ΛF )

(

)

1 iataMa ibtbMb + nFA Fa Fb

(6)

where n is the number of electrons transferred, F is Faraday’s number, A is the electrode area, i is the applied current, t is the dwell time, M is the formula weight, and F is the density. A cleaved cross section of a Pb-Tl-O modulated film was grown with layer thicknesses on the order of 0.5 µm for imaging with the SEM. Deposition was accomplished by growing an oriented prelayer followed by pulsing the potential between 70 and 230 mV vs SCE for superlattice growth. The SEM image of this modulated film is shown in Figure 9. The Pb-Tl-O film grows as a columnar structure. The superlattices

Figure 11. Polished cross section of a scanning electron micrograph of a superlattice. The superlattice maintains the two-dimensional growth around the three-dimensional defect in the sublayer. The bar marker is 10 µm.

deposited by either potential or current control all have columnar grains that grow perpendicular to the substrate. It is interesting to compare the surface morphologies of the textured prelayers to that of the superlattices. The surfaces of a [100] textured superlattice and a [210] textured superlattice are shown in Figure 10A,B, respectively. In both cases, the grains appear rounded and facets are not observed. However, the surface morphology of the [100] oriented Tl2O3 prelayer was faceted while the [210] oriented Pb-Tl-O prelayer had rounded grains (Figures 5 and 6). Figure 11 shows a polished cross section of an electrodeposited Pb-Tl-O superlattice with an individual layer thickness of 0.65 µm imaged in the SEM. A three-dimensional defect is shown in the sublayer. The alternating layer sequence is preserved around the sublayer defect.

1676 Chem. Mater., Vol. 9, No. 7, 1997

Phillips et al.

Figure 13. Comparison of the Pb-Tl-O superlattice modulation wavelength in nanometers calculated by Faraday’s law (ΛF) with those measured by scanning tunneling microscopy (ΛS) and X-ray diffraction (ΛX).

is not present in the bulk materials. These planes scatter X-rays similar to a bulk crystal, where the scattering from the ith plane can be represented by

Liλ ) 2Λ sin θi

(7)

For two planes i and j separated by a distance Λ, the path difference is:26 Figure 12. (A) [100]-textured superlattice in the Pb-Tl-O system. The modulation wavelength was calculated as ΛF ) 13.2 nm by Faraday’s law and measured by X-ray diffraction as Λx ) 13.3 nm. (B) [210]-textured superlattice with ΛF ) 12 nm and Λx ) 11.8 nm. Both superlattices were deposited by pulsing the current between 0.05 mA/cm2 (82 s) and 5 mA/ cm2 (1.2 s).

Structural evidence of a superlattice is shown by satellites which are symmetric around a Bragg reflection. Satellites result from a periodic variation in lattice spacing or composition.25 The periodic variation can be represented as a waveform. For waveforms that vary sinusoidally, first-order satellites are expected. As the waveform becomes more square, higher orders of satellites are possible. For a perfect square wave (rarely observed), only odd orders of satellites are expected. When the lattice spacing and composition both vary periodically, the intensities of like orders of satellites are not equal. The scattering factor is said to be convoluted with lattice spacing variation. An example of [210] and [100] textured superlattices grown by current pulsing are shown in Figure 12. Notice that there are first-order satellites present around each main Bragg reflection (shown in the insert boxes of Figure 12). These satellites give structural evidence that the electrodeposited films are indeed superlattices. Also, from Figure 12, unequal satellite intensities are observed in the first-order satellites for both the [100] and [210] textured superlattices. For a superlattice, the repeat distance or modulation wavelength creates an artificial interplanar spacing that (25) Guiner, A. X-ray Diffraction in Crystals, Imperfect Crystals and Amorphous Bodies, 1st ed.; W. H. Freeman & Co.: San Francisco, 1963.

(Li - Lj)λ ) 2Λ(sin θi - sin θj)

(8)

Or in terms of the modulation wavelength for X-ray diffraction, Λ ) Λx:

Λx )

(Li - Li)λ 2(sin θi - sin θj)

(9)

where Li and Lj are the satellite orders, θi and θj are the satellite positions, and λ is the wavelength of the X-ray radiation. The scanning tunneling microscope (STM) has also been used to directly measure the periodicity and modulation wavelengths of cleaved cross sections of PbTl-O superlattices.27 A comparison of modulation wavelengths measured by XRD (Λx) and STM (ΛS) to those calculated from the integrated charge (ΛF) is shown in Figure 13. The modulation wavelengths determined from the integrated charge were calculated assuming theoretical density and 100% current efficiency. The modulation wavelengths measured by XRD and STM agree for superlattice layers grown as small as 3 nm to that calculated from the integrated charge. Since the Faradaic calculations assume theoretical density, it is unlikely that the agreement would be as good if the films grew three-dimensionally instead of two-dimensionally. As the individual layer thickness increased beyond 10 nm (Λ ) 20 nm), the X-ray diffraction pattern became more complicated. Figure 14A shows the X-ray (26) Schuller, I. K. Phys. Rev. Lett. 1980, 44, 1597. (27) Golden, T. D.; Raffaelle, R. P.; Switzer, J. A. Appl. Phys. Lett. 1993, 63, 1501.

Textured Ceramic Superlattices

Chem. Mater., Vol. 9, No. 7, 1997 1677

Table 3. X-ray Powder Diffraction Results for Bulk Films Grown under Galvanostatic Control applied current density (mA/cm2)

measured potential (mV vs SCE)

lattice parametera (nm)

d spacing (hkl ) 420) (Å)

I/I0 (%) (hkl ) 420)

theor densityb (g/cm2)

0.05 1.0 5.0

70 150 230

0.5328 0.5338 0.5342

1.191 1.193 1.195

6 12 21

10.45 10.39 10.39

a Determined from the (531) reflection. b Based on the measured atomic percents from EDS and assuming all lead and thallium ions are +4 and +3, respectively.

and 5.0 mA/cm2. The Bragg peak at lower 2θ in Figure 14B has a calculated d spacing of 1.195 Å while the higher 2θ peak d spacing peak is 1.191 Å. The d spacings for the two Bragg peaks can be compared to the d spacings for individual bulk films in Table 3. The lower 2θ Bragg peak in Figure 14B corresponds to the 5.0 mA/cm2 layer, and the higher 2θ Bragg peak corresponds to the 0.05 mA/cm2 layer. This splitting of the Bragg peak may be due to a relaxation in the individual layers of the superlattice. Relaxation is expected in a strained-layer superlattice when the modulation wavelength exceeds a critical value. The lattice mismatch in this system is approximately 0.3%. The critical modulation wavelength would be expected to be smaller for superlattices with a larger mismatch. Also as the modulation wavelength increases beyond 50 nm, the peak for the 5.0 mA/cm2 layer increases in intensity compared to the peak for the 0.05 mA/cm2 layer. The higher intensity for the 5.0 mA/cm2 layer of the (420) reflection is also seen for the powder patterns of bulk films in Table 3. Our results suggest that partial relaxation of coherency occurs when the modulation wavelength exceeds 20 nm. At modulation wavelengths greater than 50 nm, the lattice parameters for the individual layers have relaxed to the values observed in bulk films grown at the respective current densities. Conclusions Figure 14. (A) X-ray diffraction pattern of a [210]-textured superlattice in the Pb-Tl-O system using Cu KR radiation as the X-ray source. The modulation wavelength was calculated as ΛF ) 50 nm by Faraday’s law. (B) Same as (A), except using Cu Kβ radiation for the X-ray source.

diffraction pattern for a Pb-Tl-O superlattice with a modulation wavelength of 50 nm. The diffraction pattern for this superlattice was obtained using the same energy wavelength (Cu KR1/Cu KR2) as in Figure 12. Notice for the superlattices in Figure 12 that KR1 and KR2 appear for each peak in the X-ray diffraction patterns. Since the energy resolution of the Si-Li solidstate detector is 150 eV and the KR1/KR2 doublet is separated by only 20 eV, the detector cannot discriminate between the KR1 and KR2 peaks. Due to the presence of this doublet in Figure 14A, it is difficult to interpret the X-ray pattern. To obtain a simpler pattern, Cu Kβ radiation was used instead to obtain the X-ray pattern for the sample in Figure 14A. Figure 14B shows the X-ray pattern for the same superlattice as in Figure 14A using Cu Kβ radiation as the X-ray source. In the X-ray pattern a splitting of the main Bragg peak occurs into two Bragg peaks with individual satellites on each side of the Bragg peak. The two Bragg peaks represent layers A and B of the superlattice grown by alternating the applied current density between 0.05

Electrodeposition was used to deposit superlattices in the Pb-Tl-O system using either current or potential pulsing. The modulation wavelengths calculated from Faraday’s law agreed well with the values determined by X-ray diffraction and scanning tunneling microscopy. The superlattices grew epitaxially and were textured perpendicular to the substrate by deposition onto either [100] oriented Tl2O3 or [210] oriented Pb-Tl-O. The prelayers and superlattices did not possess in-plane texture. Increasing texture and grain growth was observed in the prelayers perpendicular to the substrate as the films became thicker. The [100] oriented Tl2O3 prelayer films were faceted while the [210] oriented PbTl-O prelayers contained rounded grain shapes. Deposition of superlattices onto either of these prelayers resulted in grains which appear to be nonfaceted by SEM. Partial relaxation of coherency in the superlattice was observed for modulation wavelengths exceeding about 20 nm. Acknowledgment. This work was supported in part by the National Science Foundation Grant No. DMR9202872, the Office of Naval Research Grant No. N00014-96-1-0984 and N00014-94-1-0917, and the University of Missouri Research Board. CM970074I